OBJECTIVE: The objectives of this research are to evaluate the metabolic effects of dietary changes relevant to cancer and to assess the safety, pharmacokinetics, bioavailability, and mechanisms of action of macro- and micronutrients with cancer chemoprevention potential. BACKGROUND: Clinical nutrition studies provide information that helps bridge the gap between observational research and clinical trials by assessing potential mechanisms of action and other parameters important in developing intervention strategies. METHODS: Since 1983, collaborative efforts with the Beltsville Human Nutrition Research Center of the USDA in clinical nutrition research has resulted in conduct of over 10 studies, focused primarily on antioxidant and hormone research. Studies have evaluated the dose, bioavailability, and safety of selenium, carotenoids, and vitamin C, and the potential modulating role of dietary fat and alcohol on hormones. PROGRESS: PAST STUDIES - (1) Evaluation of carotenoids in premenopausal women on and off alcohol on the same controlled diet showed that plasma levels of both beta- and alpha-carotene were higher on alcohol, while lutean/zeaxanthin levels were lower. (2) In another controlled diet study, all plasma carotenoid levels were found to vary over the menstrual cycle: concentrations were lowest at menses and each peaked following the peak of its lipoprotein carrier. (3) Hormone evaluations in premenopausal women have shown that taller women have higher estradiol (E2), that pregnancy may modify the age-related changes in E2 levels (E2 decreases with age in parous women but increases in nulliparous), that energy intake is inversely related to serum levels of androstenedione and DHEAS, and P:S ratio is inversely related to E2 and E1 levels. (4) In men, a controlled high-fat/low-fiber diet resulted in higher plasma levels of total and SHBG-bound testosterone, higher urinary testosterone, and lower urinary estrogens compared to a low-fat/high-fiber diet. RECENT STUDIES - (1) A recently completed controlled alcohol feeding study in postmenopausal women determined that alcohol ingestion led to increased levels of both estrone sulfate and DHEA sulfate, providing support for one possible mechanism of action for the observed relation of alcohol to breast cancer risk. In other analyses from this study, alcohol consumption reduced lipid-related risk factors for cardiovascular diseases, bioavailable IGF-1, and serum B12 levels; improved insulin sensitivity; increased serum leptin and some markers of oxidative stress; and did not affect autoantibodies to DNA damage. Dietary methodology studies conducted as part of this study compared energy expenditure from four physical activity questionnaires with doubly labeled water estimates, and documented calorie intake misreporting by diet record and food frequency questionnaires (also in relation to doubly labeled water). (2) A study of the pharmacokinetics of selenium, a particularly promising chemopreventive agent, has been completed and statistical modeling is ongoing.